Join Bridge Winners
Occam's Razor
(Page of 6)

Near the start of the second quarter of a close Rosenblum match, you get a chance to act over partner's multi.

N-S vul, East deals. As West, you hold:

West
KJ10952
QJ6
108
J3
W
N
E
S
2
X
?

2: Weak 2 in one of the majors. You play a very liberal style of multi. Partner could easily have a 5-card suit -- in fact, at this vulnerability he will have a 5-card suit far more often than a 6-card suit. The suit quality does not have to be particularly strong, and he might or might not have cards on the side.

DBL: Either a takeout double of spades or some strong hand.

As partner of the multi bidder, you are in complete control of the auction. You have many options available:

Pass: If 4th seat bids, partner shuts up. If 4th seat passes, partner passes (ending the auction) if he holds 3+ diamonds but he shows his suit with fewer than 3 diamonds. Not applicable here, but on hands with 4+ diamonds and short in a major this will often avoid going for a number since 4th seat needs a diamond stack to pass.

Redouble: Forces partner to bid 2, after which you place the contract. This allows you to stop in 2 when you have a long heart suit.

2, 2, 3, 3: These are all pass or correct bids. Partner passes if you hit his suit, but bids if you don't hit his suit.

2NT: Asking about partner's hand. At least game-invitational.

3: Shows hearts, asks about heart length. At least game invitational.

3: Shows spades, asks about spade length. At least game invitational.

3NT, 4, 4: End of auction.

4: Asks partner to transfer to his suit.

4: Asks partner to bid his suit.

Which do you choose?


West
KJ10952
QJ6
108
J3
W
N
E
S
2
X
?

If partner had opened a weak 2 bid, you would raise to 3. You are willing to take your chances with at least an 8-card fit and the opponents flying blind. Therefore, you do so here. It is conceivable that partner has spades, and if he does he will not pass 3 and you can continue accordingly.

The bidding continues:

W
N
E
S
2
X
3
P
P
3NT
?


Are you done?:



West
KJ10952
QJ6
108
J3
W
N
E
S
2
X
3
P
P
3NT
?


You are done. 4 could easily go for too much, and you might beat 3NT. You have done as much damage as you can to the opponents.

If there were any doubt about your pass, you should be deciding on what you would lead before you pass. Your chances of defeating the contract might determine whether you should bid, pass, or double, and you can't make a good assessment of these chances until you have chosen your opening lead if you decide to defend. On this hand the pass is very clear regardless of what you will lead.

That ends the auction.

W
N
E
S
2
X
3
P
P
3NT
P
P
P

Your lead. Your lead methods vs. notrump are:

From an honor sequence, you lead ace from ace-king, king being your power lead asking for count or unblock, and Rusinow (second high from touching honors) with 3 or more cards in the suit for lower sequences. However, in partner's suit you lead standard. Thus, if you choose to lead a heart your expected lead would be the queen, while if you choose to lead a spade honor your expected lead would be the 10.

If you are not leading an honor (9 or higher is considered an honor), you are expected to lead the lowest card if you wish to show strength in the suit, the highest spot you can afford if you do not wish to show strength in the suit.

Assuming you choose to lead a spade, which spade would you lead?



West
KJ10952
QJ6
108
J3
W
N
E
S
2
X
3
P
P
3NT
P
P
P


If you had a minor-suit ace, you would give serious consideration to the expert king of spades lead, hoping to pin somebody's stiff queen. That will get you into the newspapers when it works, but in practice it seems to either not matter or backfire. With no minor-suit entry, leading the king is clearly wrong.

Your proper spade lead is the 10 of spades. But there is no law which says you must do what is proper. If the information from your lead will be more valuable to declarer than to partner, it may be correct to make a non-standard lead.

Can it matter to partner what you lead? If he has the ace or the queen he will play it and return the suit regardless of which spade you lead. It clearly won't make a difference to partner.

Can it matter to declarer? It might. The relevant layout is dummy having Qxx, partner ace-doubleton, and declarer two small. If declarer goes up queen you run the suit, but if declarer ducks the suit is blocked.

Which lead is most likely to persuade declarer to play the queen? If you lead the 10, declarer probably won't play you for ace-king, since with AKJ10 or AK109 you would likely have led one of the honors to take a look -- you would be worried about losing to a doubleton queen. However, with AK9xx(x) you might well lead low, hoping it is partner who has the doubleton queen. Therefore, your best chance to persuade declarer to fly with the queen is to lead a low spade rather than one of your honors.

Can leading low cost? Not really. If dummy has Q8x and declarer has a singleton (possible since the double might have been a takeout double of spades), leading low would cost if declarer could look into your hand. In practice, declarer will always go up with the queen if that is the layout. Also leading low might lose to the 8-spot, but if that happens a different spade lead wouldn't have been successful anyway.

Does it matter which spade spot you lead? Probably not, but it might. Your best bet is to lead the 5 of spades. Declarer will be expecting you to be leading your lowest spade from a good suit. By concealing the 2, you make it look more likely that you have a 5-card spade suit, since declarer will think your partner has the 2. If declarer has Qxx opposite a doubleton he will definitely go up queen if he believes you have only 5 spades, while if he thinks you have 6 spades he might duck to block the suit. Also, suppose partner has Q4 or Q3 doubleton. Partner will win the queen and return the suit. If declarer can see that partner is returning the lowest spot he might not hold up the second round, since he can expect your partner would have returned the higher of his remaining doubleton. However, if you conceal the 2 declarer will believe that partner has that card and will hold up the second round, which is what you want to see happen.

Now, on to the bigger question. Do you lead a spade or a heart?



West
KJ10952
QJ6
108
J3
W
N
E
S
2
X
3
P
P
3NT
P
P
P


The spade lead will probably succeed if partner has the ace of spades, since when dummy has Qxx you don't expect declarer to get it right and duck. Only when partner's ace is singleton or one of the opponents has Qxxx of spades will the spade lead not work when you hit partner's ace. The spade lead also has a small chance when partner has the queen of spades, either because declarer isn't able to hold up long enough or because after he ducks two rounds you can shift to a heart and establish the setting trick.

The heart lead needs a parlay. Declarer definitely has a heart stopper for his 3NT call, while he might not have a spade stopper. You need for partner to have good enough hearts that declarer has only one stopper. If your multi were more disciplined that would be very likely, but in your style at this vulnerability partner could easily have K9xxx or 109xxxx. You need partner to have a minor-suit winner. In addition, that minor-suit winner must score before declarer can cash 9 tricks.

There is a principle in science called Occam's Razor. The idea is that when you have a choice of possible explanations for an observation, you should choose the explanation which requires the fewest number of assumptions. The same principle can apply in bridge, particularly with opening leads. Make the lead based on the fewest number of assumptions needed to defeat the contract.

This hand is a good example. Clearly any one of the 3 things needed for the heart lead to be right (good enough heart suit, minor-suit entry, declarer not having 9 cashers) is more likely than partner having the ace of spades. But when you multiply all the probabilities of these chances together, the result doesn't appear to be as high as the single chance of partner having the ace of spades. occam's Razor says the spade lead is the percentage choice, as it requires the fewest assumptions.

What about the fact that your partner bid hearts. Is there a danger of hurting partnership morale if you don't lead his suit when that would have been the winner? Some players feel that way and will always lead partner's suit if the decision looks close, since they think that if they are wrong at least they have done what partner suggested. In my mind, this is nonsense. The only lead which hurts partnership morale is the losing lead. Partner bidding a suit is far from a command to lead the suit. It is simply one of many pieces of information for you to process to find the best lead.

You choose to lead the 10 of spades. This is not a success, as the full hand is:

West
KJ10952
QJ6
108
J3
North
Q87
2
Q97654
Q74
East
6
K10987
32
A10852
South
A43
A543
AKJ
K96
W
N
E
S
2
X
3
P
P
3NT
P
P
P
D
3NT South
NS: 0 EW: 0
10
7
6
A
3
1
0
A
8
4
2
3
2
0
6
3
Q
A
2
2
1
3


Declarer played low from dummy (obviously the safest play for the contract), won the ace, tested the diamonds to make sure they weren't 4-0, knocked out the ace of clubs, and had 9 tricks. On the heart lead he would have had no chance.

I was the West defender who found the unfortunate spade lead. Obviously it was a disaster on this hand. Most players disagree with my choice. I think it is a close decision, but on the information I had I still believe my choice of a spade is the percentage action.

Why did I lead the 10 rather than the superior 5? Very simple. I never thought about the advantage of leading low. There was an error in my thinking which illustrates a very common methodological mistake.

As an example, let's suppose you are considering whether or not to pass or make a game try. The proper methodology is to first analyze which game try is best if you choose to make the game try. Having made that decision, you are now ready to make the more important decision of whether to make the game try. There are two reasons. First, having chosen the best game try you are better placed to evaluate the effect of the game try vs. passing. Second, if you do make the game try you will be making the right bid. However, since it is obvious that the big decision is whether or not to make a move, many players will wrongly focus on that decision first and possibly not find the best game try.

This lead problem is an illustration. My proper thinking should first be to decide which card I will lead from each suit if I choose to lead the suit (obviously I will be leading a major here, so I don't have to worry about the minors, and the queen of hearts is an easy choice if I choose to lead a heart). Once I have made that decision I can then focus on the bigger question of which suit to lead, and I will be leading the right card in the suit and making the evaluation based on the proper lead. At the table I focused only on the choice of which suit to lead, and never realized that the choice of which spade to lead could make a difference. My only consolation was that had I led a low spade declarer would have made an overtrick.

What do you think of the philosophy of opening multi (or a weak 2 if not playing multi) on partner's hand?



West
KJ10952
QJ6
108
J3
North
Q87
2
Q97654
Q74
East
6
K10987
32
A10852
South
A43
A543
AKJ
K96
W
N
E
S
2
X
3
P
P
3NT
P
P
P
D
3NT South
NS: 0 EW: 0
10
7
6
A
3
1
0
A
8
4
2
3
2
0
6
3
Q
A
2
2
1
3


There are pros and cons to such a bid. Some of the cons are:

1) You might walk into a misfit and go for a number. This isn't very likely. You have a decent suit, and at this vulnerability the opponents will have to be on very sure ground to go for a penalty vs. 2 since they may have a vulnerable game so the penalty might not be sufficient compensation.

2) You might be playing in the wrong strain. This is a real danger. You could have a big club fit, and it will be difficult to find this fit after preempting in hearts. Of course it might not be easy to find the fit if you pass.

3) Partner might misjudge whether to bid a game, which game to bid, or how high to compete. This is always a danger with an offbeat preempt. Still, passing isn't exactly a perfect description of the East hand. It is true that by passing East might be able to make a more descriptive call with Michaels or unusual notrump, but the auction won't always time out properly for that. In the meantime, the opponents will have had the opportunity to exchange valuable information at a low level.

Some of the pros are:

1) First strike. The pair who gets the first bid in is generally at an advantage. They have made some kind of descriptive call, while their opponents have to start from scratch without the benefit of carefully discussed constructive auctions.

2) Preemptive value. The opponents have to start bidding at a high level. While sometimes your preemption benefits the opponents, more often it causes them to have some kind of accident.

3) Getting in quickly. Quite often it is safer to preempt than to pass. If you pass, you may have to guess whether or not to compete later. If you sell out too cheaply, that can be costly. But if you pass and then enter the auction it will be easier for the opponents to double when the hand is a misfit since they will have exchanged some information.

4) Hitting a fit. If the preempt hits partner with a fit, he can compete quickly. Sometimes he can bid a game which might not have been biddable without the preempt. Other times he can jack the bidding up making life difficult for the opponents. The actual hand is a good example. South pretty much had to bid 3NT, but it could have been a ridiculous contract down several if he caught a bad dummy.

5) Lead value. If partner is the opening leader, the preempt may help him. This hand is a good example. If East had never bid, it would be very difficult for West to find the heart lead.

Most players underestimate the importance of vulnerability when it comes to preempts. They understand that they can be more aggressive non-vulnerable, particularly if the opponents are vulnerable, since if they go for a number that number is less than if they were vulnerable and the opponents may have a vulnerable game available. That is only a small part of the picture. Let's say you make some offbeat preempt at favorable vulnerability. If something bad happens (you go for 800 against their game, you go for 300 or 500 against their part-score, you are too high on a misfit, you play the wrong part-score, you bid a bad game, or you miss a good game), the maximum loss is generally around 6 IMPs. But when it is the opponents who have the accident (they miss a good game, they play the wrong game, they go for a number, or they make a wrong slam decision), the gain is usually in double figures. Most players will be aggressive bidding vulnerable games, since they are getting 10 to 6 odds. Yet, these players don't realize that they are getting the same 10 to 6 odds with very aggressive preempts at favorable vulnerability.

Since our preempts can be quite normal as well as far out, players often ask me how responder can handle this. My answer is that responder doesn't worry about it. He assumes the preempt is somewhere between normal and far out, and makes his best guess. He might guess wrong, but at least he has some idea of what his partner has. The opponents are more likely to guess wrong, since they start with no knowledge about their partner's hand. Sure, those who play sound preempts will generally do better when they are dealt a perfect preempt, although their opponents might benefit just as much in the play of the hand. However, we get to preempt a lot more often than the sound preempters, and this more than makes up for responder's problems.

19 Comments
Getting Comments... loading...
.

Bottom Home Top